Overestimation Unveiled: Men’s Skill Overrating and the Power of Question Phrasing to Counteract It

Overestimation bias is a critical issue in subjective assessments in survey research, with certain sociodemographic groups, such as men, being more prone to overestimate their schooling, earnings, or skills. This paper investigates how overestimation bias arises in self-assessments of whether individuals’ skills match their job requirements and explores the conditions under which this bias manifests. I conducted an online survey with a split-half design, enabling a comparative analysis of two instruments for measuring subjective skill mismatches. Both instruments require a self-assessment of whether an individual’s skills exceed, match, or fall short of job requirements, with version 1 slightly emphasizing individual skills and version 2 emphasizing job requirements.

The study reveals statistically significant differences between the two versions. Version 1 shows roughly twice the incidence of overskilling, lower rates of skill matching, and half the incidence of underskilling compared to version 2. The skill mismatch distributions in version 2 are more balanced, with predictive analyses indicating higher construct validity for this version.

These discrepancies also vary across sociodemographic groups. Notably, the differences between versions are more pronounced for men than for women. While women show similar overskilling rates across both versions, men assess themselves as overskilled more than twice as often in version 1. This suggests that men are more likely to be affected by overestimation bias, particularly when the assessment emphasizes their own skills.

My findings enhance the understanding of biases in self-reports, demonstrating how question phrasing influences overestimation bias and how slight adjustments can mitigate this issue. Additionally, our research shows that overestimation of skills is predominantly driven by men. The implications of this research are crucial for understanding, interpreting, and addressing gender inequalities in the labour market. By revealing how emphasis in self-assessment affects biases, this study underscores the need for precise survey design.